Chaitanya Ahuja

MTS @AIatDoorDash | past: @AIatMeta, PhD @SCSatCMU | AI researcher

[Chaitanya Ahuja @chahuja]

@TheOneAndArjun @aryanxshah @andyfang DM should be open now

[Chaitanya Ahuja @chahuja]

Personal Update: Excited to share that I've joined DoorDash's newly minted AI Research team as a Member of Technical Staff, working alongside @aryanxshah and @andyfang. We're defining the next era of post-training for agentic commerce, memory, multimodal understanding, and robotics for last-mile delivery.

We're hiring! DM me if you want to help shape our research agenda, technical direction, and mission.

The last four years at Meta were an incredible adventure, and I'm deeply grateful to the colleagues who I had the opportunity to work with.

This also comes with a move to the San Francisco Bay Area. Looking forward to reconnecting with old friends and colleagues out here.

[Chaitanya Ahuja @chahuja]

C-DiffGAN encourages knowledge retention by 1) generating reminiscences of previous low-resource speaker data, then 2) crossmodally aligning to them to mitigate catastrophic forgetting

5/n

[Chaitanya Ahuja @chahuja]

Adapting Generative models for co-speech gestures to new speakers, but without forgetting the style of previous speakers? Let’s make it more challenging by only having 2-10 minutes of data for the new speakers.

Website: https://t.co/iJI9W5GXId

Find us at #ICCV2023

1/n

[Chaitanya Ahuja @chahuja]

We propose an approach, C-DiffGAN, to learn a multi-speaker co-speech gesture generation model in a continual learning setting. i.e. By watching only 2 minutes of videos for every speaker in a sequence (without having access to previous speakers’ data).

4/n

[Chaitanya Ahuja @chahuja]

Challenging and enlightening as it was, it wouldn't have been possible without my wonderful co-authors Simbarashe Nyatsanga, @SvitozarTaras, Gustav Henter, and Michael Neff

[Chaitanya Ahuja @chahuja]

Last year we surveyed the body of co-speech gesture generation work with a heavier focus on the past decade of data-driven learning.

Link: https://t.co/ThcsCHSW5a

[Dmytro Mishkin @ ECCV2026 @ducha_aiki]

A Comprehensive Review of Data-Driven Co-Speech Gesture Generation

Simbarashe Nyatsanga,@SvitozarTaras , @chahuja, Gustav Eje Henter, Michael Neff

tl;dr: if you are starting to work in gesture recognition/generation, that is a good read to start.

https://t.co/PYFrueB6UU

[Jayesh K. Gupta @rejuvyesh]

We're hiring interns in the Autonomous Systems and Robotics Research group here at Microsoft.
If you are a PhD student interested in working at the intersection of fluids simulations and ML, please consider applying:

https://t.co/HeFXV8Uf9p

[Chaitanya Ahuja @chahuja]

@bhuwandhingra Congratulations Bhuwan!!

[Chaitanya Ahuja @chahuja]

@LTIatCMU @mldcmu
5/n

[Chaitanya Ahuja @chahuja]

Learning to generate personalized co-speech gestures from spoken language, but do not have 5-10 hours of supervised training data for that person?

Check us out today at #CVPR2022 poster session 4.2 from 230-500pm at poster 180b
w/ @_dongwonlee @lpmorency
1/n

[Chaitanya Ahuja @chahuja]

More co-speech gesture generation work
1 https://t.co/kIwx6e9d9P
2 https://t.co/Ahwu88IGe6
3 https://t.co/IC7exQhPoV
4 https://t.co/IBPnx3RSQy
5 https://t.co/iSAymZF3UX
6 https://t.co/DRoZe3kZhl
7 and many more

DM me if you are starting out in this area or just want to chat
4/n

[Chaitanya Ahuja @chahuja]

Check out this workshop at #CVPR2022 (and upcoming survey paper). It beautifully builds a taxonomy encompassing key challenges of multimodal ML research

[Paul Liang @pliang279]

If recent models like DALL.E, Imagen, CLIP, and Flamingo have you excited, check out our upcoming #CVPR2022 tutorial on Multimodal Machine Learning - next monday 6/20 9am-1230pm

https://t.co/nnBbipg34G

slides, videos & a new survey paper will be posted soon after the tutorial!

[Chaitanya Ahuja @chahuja]

@SurbhiGoel_ @PennCIS Congratulations Surbhi!!

[Paul Liang @pliang279]

Really excited to release the video of my guest lecture on Multimodal Deep Learning for CMU's Deep Learning class @rsalakhu @mldcmu
It covers 5 fundamental concepts in multimodal ML: representation, alignment, reasoning, translation & co-learning

Youtube: https://t.co/I9Ukv0r2NE

[Zachary Lipton @zacharylipton]

Tomorrow, we launch "The Art of the Paper", a course on the principles, mechanics, & culture of scientific writing. While core to our work, this material seldom gets a formal treatment. I'm excited, nervous, & grateful to @mldcmu for the creative freedom.

https://t.co/iHtTVV469X